This study was approved by the local ethics committee of RWTH Aachen University (CTC-A-No.: 21-083; Ethics-No.: EK 092/21), registered on ClinicalTrials.gov (ID NCT06061120), and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all subjects. The authors had full access to the data and declare that all supporting data are available in the article. They accept responsibility for the data's integrity and agree to the article as written.

The trial was a prospective, hypothesis-generating, single-arm, single-center cohort study. Cryoballoon ablations and ECG recordings were performed in 2021 and 2022 at the University Hospital RWTH Aachen. A total of 69 subjects were enrolled in the study. Inclusion criteria were: Documented atrial fibrillation with subsequent planned cryoballoon PVI, ECG recording using YRS-100 device (YourRhythmics, Maastricht, The Netherlands), the ability and willingness to provide informed consent, and an age of at least 18 years. Exclusion criteria were: Previous PVI, non-cardiovertable rhythm other than sinus rhythm during ablation, emergency ablation and unstable condition before ablation.

The study population is visualized in Figure 1. A total of 69 subjects were initially enrolled. Three subjects were excluded due to a non-cardiovertable rhythm other than sinus rhythm at the begin of the ablation procedure.

Individual ECG recordings were excluded if a non-sinus rhythm was present or if P-wave detection by bespoke software was insufficient due to poor ECG recording quality or external pacing. Individual leads were excluded in cases of insufficient signal quality or electrode disconnection.

Of the remaining 66 subjects, 10 had at least one ECG recording excluded, and 22 subjects had at least one lead excluded from an ECG recording.

ECG recordings were performed with a sampling rate of 2,000 Hz and a duration of at least 5 min per ECG recording. A total of 22 leads were recorded. Of these, 12 leads corresponded to those of a standard 12-lead ECG. Based on prior studies (26–31) 10 additional unipolar leads (×01–10; Figure 2) were included. This lead placement has been shown to maximize derived atrial electrical potential differences (26–31). Landmarks were assigned to the electrode positions to allow for standardization (Figure 2). For the additional lead positions, these included the fifth intercostal space at the right and left mid-axillary line, the manubrium of the sternum, the xiphoid process and the vertebra prominens.

The study protocol was structured as shown in Figure 3. ECG recordings were taken immediately before the start of the procedure (baseline) and after ablation of each PV. The final available recording following the last ablated PV per patient is referred to as the post procedure measurement. The sequence of ablation was at discretion of the treating physician (Supplementary Table S1). Left (LPVs) and right pulmonary veins (RPVs) were always ablated sequentially, typically starting with the LPVs and the left superior pulmonary vein (LSPV). Once the cryoballoon reached a temperature of 20 °C after ablation with subsequent deflation, the ECG recording was started. No manipulation was performed on the subject during the recording. Demonstrated elimination of all PV potentials using a lasso catheter was the endpoint for PVI. The P-waves were compared between baseline and post procedure ECG recordings, between before and after ablation of each PV, and between before and after ablation of LPVs or RPVs. If a left or right common ostium was present, its ablation was included in the comparison of LPVs or RPVs but excluded from the comparison of individual PVs on that side. Subjects with a heart rhythm other than sinus rhythm were electrically cardioverted if possible or excluded from the study.

IBM SPSS Statistics software (IBM Corp. Released 2021. Version 28.0. Armonk, NY, United States) was used for statistical analysis. The Shapiro–Wilk test was used to assess the normality of distribution. For normally distributed continuous variables, a two-tailed paired t-test was applied. For non-normally distributed data, the Wilcoxon signed-rank test was used. In Table 2 and throughout the manuscript, a # symbol indicates parameters analyzed with the Wilcoxon signed-rank test. The P-wave terminal force was considered only in lead V1. All other parameters were calculated averaged across all leads. Continuous variables are presented as mean ± standard deviation to ensure consistency and facilitate comparison, even when non-parametric tests were applied. Categorical variables are reported as number and percentage. A p-value < 0.05 was considered statistically significant.

Table 1 summarizes the baseline characteristics of the participants. Most subjects were male (70%) with an average age of 64 ± 10 years. The mean body mass index was 29.1 ± 5.3 kg/m² and the mean CHA2DS2-VA score was 1.9 ± 1.3. While 70% of patients had paroxysmal atrial fibrillation, 79% were in sinus rhythm at the start of the procedure. The average left ventricular ejection fraction was 54 ± 6%, the left atrial diameter was 41 ± 5 mm, and the left atrial area was 22.8 ± 6.3 cm². Most patients were treated with beta-blockers (83%) and oral anticoagulants (94%).

Changes in all ECG parameters are shown in Table 2 and Figure 5. The signal-averaged P-waves at baseline and after ablation of each PV for a sample subject are visualized in Figure 6.

Currently, procedural success after PVI is typically assessed by monitoring for recurrence of atrial fibrillation. However, this endpoint becomes evident only after a longer observation period and depends on the occurrence of clinical arrhythmia episodes. In contrast, high-resolution signal-averaged ECG may provide insights into ablation-related electrophysiological changes already during sinus rhythm, offering complementary information at an earlier stage and independently of arrhythmia recurrence.

While electroanatomical mapping and cardiac MRI provide superior spatial and mechanistic insights, they are time-consuming, costly, and typically not performed in single-shot ablation procedures. In contrast, high-resolution signal-averaged ECG represents a non-invasive, cost-efficient, and rapidly applicable alternative that may be feasible for repeated peri- and postprocedural use in routine clinical practice.

A key challenge of this study was the electrocardiographic visualization of subtle changes in the left atrium following ablation of individual PVs. To optimize left atrial visualization, additional lead positions were obtained from the front and back based on preliminary studies (26–31). A high-resolution ECG with a sampling rate of 2,000 Hz and a prolonged recording time of 5 min enabled the calculation of a signal-averaged P-wave. While signal averaging substantially improves the signal-to-noise ratio and allows analysis without heavy filtering, it inevitably suppresses beat-to-beat variability and may smooth subtle morphological features. To be independent of the investigator, custom software was used to determine the P-wave onset and offset, as described in the methods section and in previous work (26).

Beyond its potential procedural applications, this concept also relates to the broader framework of atrial cardiomyopathy, which encompasses structural, electrical, and functional remodeling processes predisposing to atrial fibrillation (47). While the present study focused on acute iatrogenic substrate modification during ablation, the same methodological approach could also be applied to study electrophysiological manifestations of chronic pathological remodeling. High-resolution signal-averaged ECGs could therefore represent a promising non-invasive approach to characterize atrial substrate changes over time, whether induced by disease or therapeutic intervention.

To the best of our knowledge, the present study is the first to describe the stepwise electrocardiographic changes in the P-wave during cryoballoon PVI of individual PVs. We found that ablation of the LPVs, particularly the LSPV, was followed by measurable changes in P-wave parameters. Intraprocedural changes in P-wave terminal force and aspects of P-wave morphology were consistent with patterns reported in previous studies analyzing ECG alterations after completed PVI (19, 22–24). While these alterations may be related to short-term electrophysiological remodeling, further research is needed to determine their persistence over time and their potential relevance for evaluating ablation efficacy and predicting long-term rhythm outcomes.

Due to the limited sample size, this study should be regarded as exploratory and hypothesis-generating. Nevertheless, consistent trends in ECG changes were observed, suggesting that larger studies may help confirm and further refine these preliminary findings.

The sequence of PVI was not randomized but followed routine procedural practice, as outlined in Supplementary Table S1. While this reflects common clinical workflows, it may limit the ability to isolate vein-specific effects with certainty. In addition, cumulative effects from preceding ablation steps on later-ablated PVs cannot be excluded. Therefore, the reported changes cannot be conclusively attributed to isolated vein-specific effects.

Intraprocedural ECG recordings may be influenced by transient procedural factors such as edema formation (35, 36) or autonomic responses (37–44). However, these effects are expected to occur in all patients and were mitigated by a standardized protocol: Recordings began only after the cryoballoon reached ≥20 °C and were conducted without catheter manipulation over a continuous 5-minute period. Given the consistent cryoablation technique, observed changes likely reflect a reproducible combination of general procedural effects and localized electrophysiological responses.

Detailed procedural variables such as time-to-isolation or quantitative PV potential characteristics were not consistently available across all veins and were therefore not included. Long-term follow-up was not part of the predefined analysis, as the study focused exclusively on acute intra-procedural surface-ECG changes.